Avoid or eliminate duplicated/reverse results in a dictionary in Python

Question

enter image description here

I am looking for pairs in two columns in a csv. If it finds AB and BA include just AB in the dictionary. I wrote the following script:

dataset = list(zip(col1,col2))
for a,b in dataset:
    if (a,b) and (b,a) in dataset:
        dic [a] = b 

But obviously I only need one and the output is:

{'A': 'B', 'C': 'B', 'B': 'A', 'D': 'C', 'F': 'C', 'H': 'C', 'J': 'X', 'X': 'J'}

As you see it is duplicated sometimes (for example, D:C is correct but it only appears once, I don't know why)

How can I avoid those duplicates or eliminate from the dictionary the "reverse forms"?

Answer 1

Try a simple addition:

if (a,b) and (b,a) in dataset \
    and a < b :

This gets you only one of the two. It also assumes (as in your example) that there is no row containing the same value twice.

Answer 2

You don't have to check both pairs in each step. Instead you only need to check if b is in the dictionary and if so, whether dic[b] != a .

The reason for this is that we are always adding from col1 as the key, so we only need to see if the reversed value has already been added.

col1 = ['A', 'C', 'B', 'D', 'X', 'F', 'H', 'J']
col2 = ['B', 'B', 'A', 'C', 'J', 'C', 'C', 'X']

dic = {}
for a, b in zip(col1, col2):
    if (b not in dic) or (dic[b] != a):
        dic[a] = b

#{'A': 'B', 'C': 'B', 'D': 'C', 'F': 'C', 'H': 'C', 'X': 'J'}

However if you only wanted to keep one copy of each pair where both versions existed, you would need a slightly different approach.

First create a dictionary with all of the common pairs. Then iterate as before, and only add if both versions exist. Use the same logic as above to avoid duplicates.

d = dict(zip(col1, col2))
dic = {}
for a, b in d.items():
    if (a in d) and (b in d) and (d[a] == b) and (d[b] == a):
        if (b not in dic) or (dic[b] != a):
            dic[a] = b

#{'A': 'B', 'J': 'X'}